Chris Nennemann
6771a4c235
Python CLI tool that fetches AI/agent-related Internet-Drafts from the IETF Datatracker, rates them using Claude, generates embeddings via Ollama for similarity/clustering, and produces markdown reports. Features: - Fetch drafts by keyword from Datatracker API with full text download - Batch analysis with Claude (token-optimized, responses cached in SQLite) - Embedding-based similarity search and overlap cluster detection - Reports: overview, landscape by category, overlap clusters, weekly digest - SQLite with FTS5 for full-text search across 260 tracked drafts Initial analysis of 260 drafts reveals OAuth agent auth (13 drafts) and agent gateway/collaboration (10 drafts) as the most crowded clusters, while AI safety/alignment is underserved with the highest quality scores. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
153 lines
7.8 KiB
Markdown
153 lines
7.8 KiB
Markdown
# IETF AI/Agent Draft Overlap Analysis
|
|
*Generated 2026-02-28 — 260 drafts analyzed, embeddings via nomic-embed-text*
|
|
|
|
## Summary
|
|
|
|
- **42 clusters** at 0.85 similarity threshold (topically overlapping)
|
|
- **34 clusters** at 0.90 threshold (near-duplicates or same-author variants)
|
|
- Heaviest overlap in OAuth agent auth (13 drafts) and agent gateway/collaboration (10 drafts)
|
|
|
|
---
|
|
|
|
## Near-Duplicates (same draft, different WG or renamed)
|
|
|
|
| Draft A | Draft B | Reason |
|
|
|---------|---------|--------|
|
|
| draft-a2a-moqt-transport | draft-nandakumar-a2a-moqt-transport | Same draft |
|
|
| draft-abbey-scim-agent-extension | draft-scim-agent-extension | Same draft |
|
|
| draft-ahn-nmrg-5g-security-i2nsf-framework | draft-ahn-opsawg-5g-security-i2nsf-framework | Different WG |
|
|
| draft-ar-emu-hybrid-pqc-eapaka | draft-ietf-emu-hybrid-pqc-eapaka | WG adoption |
|
|
| draft-aylward-aiga-1 | draft-aylward-aiga-2 | Version |
|
|
| draft-bastian-jose-dvs | draft-bastian-jose-pkdh | Renamed |
|
|
| draft-bernardos-cats-isac-uc | draft-bernardos-green-isac-uc | Different WG |
|
|
| draft-cui-nmrg-llm-nm | draft-irtf-nmrg-llm-nm | WG adoption |
|
|
| draft-eggert-mailmaint-uaautoconf | draft-ietf-mailmaint-pacc | WG adoption |
|
|
| draft-happel-structured-email-trust | draft-ietf-sml-trust | WG adoption |
|
|
| draft-ietf-httpbis-layered-cookies | draft-ietf-httpbis-rfc6265bis | Versioned |
|
|
| draft-ietf-lamps-est-renewal-info | draft-yusef-lamps-rfc7030-renewal-recommendation | Same topic |
|
|
| draft-lake-pocero-authkem-edhoc | draft-pocero-authkem-edhoc | Renamed |
|
|
| draft-li-dmsc-macp | draft-li-dmsc-mcps-agw | Evolution |
|
|
| draft-men-rtgwg-agent-networking-digibank-scenario | draft-men-rtgwg-agent-networking-in-digibank | Renamed |
|
|
| draft-pang-agents-networking-scenarios | draft-zl-agents-networking-scenarios | Same content |
|
|
| draft-rosenberg-aiproto | draft-rosenberg-aiproto-nact | Renamed |
|
|
| draft-rosenberg-aiproto-cheq | draft-rosenberg-cheq | Renamed |
|
|
| draft-sun-zhang-iaip | draft-sz-dmsc-iaip | Different WG |
|
|
| draft-templin-manet-inet | draft-templin-manet-inet-omni | Evolution |
|
|
| draft-tong-network-agent-use-cases-in-6g | draft-yu-ai-agent-use-cases-in-6g | Competing |
|
|
| draft-zeng-mcp-troubleshooting | draft-zm-rtgwg-mcp-troubleshooting | Different WG |
|
|
| draft-zheng-agent-identity-management | draft-zheng-dispatch-agent-identity-management | Different WG |
|
|
| draft-zhul-dhc-bnc-up-specific-suboption | draft-zhul-intarea-bnc-up-specific-suboption | Different WG |
|
|
| draft-zl-agents-networking-architecture | draft-zl-agents-networking-framework | Same series |
|
|
|
|
---
|
|
|
|
## Major Competing Clusters
|
|
|
|
### 1. OAuth for AI Agents (13 drafts) — MOST CROWDED
|
|
|
|
All trying to solve: how do AI agents authenticate and get authorized via OAuth?
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-aap-oauth-profile | OAuth 2.0 profile for autonomous agents |
|
|
| draft-aylward-daap-v2 | Distributed accountability protocol |
|
|
| draft-barney-caam | Contextual agent authorization mesh |
|
|
| draft-chen-ai-agent-auth-new-requirements | New auth requirements analysis |
|
|
| draft-chen-oauth-rar-agent-extensions | RAR extensions for agent policy |
|
|
| draft-goswami-agentic-jwt | Agentic JWT for autonomous systems |
|
|
| draft-jia-oauth-scope-aggregation | Scope aggregation for multi-step workflows |
|
|
| draft-liu-agent-operation-authorization | Verifiable delegation via JWT |
|
|
| draft-liu-oauth-a2a-profile | A2A profile for transaction tokens |
|
|
| draft-oauth-ai-agents-on-behalf-of-user | On-behalf-of-user extension |
|
|
| draft-song-oauth-ai-agent-authorization | Target-based authorization |
|
|
| draft-song-oauth-ai-agent-collaborate-authz | Multi-agent collaboration authz |
|
|
| draft-yao-agent-auth-considerations | Auth considerations analysis |
|
|
|
|
**Gap:** Most focus on single-agent auth. Few address chained delegation across multiple agents or revocation in real-time agent-to-agent workflows.
|
|
|
|
### 2. Agent Gateway / Multi-Agent Collaboration (10 drafts)
|
|
|
|
All trying to solve: how do agents from different platforms/ecosystems collaborate?
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-agent-gw | Semantic routing gateway |
|
|
| draft-campbell-agentic-http | HTTP best practices |
|
|
| draft-cui-dmsc-agent-cdi | Cross-domain interop framework |
|
|
| draft-fu-nmop-agent-communication-framework | Network AIOps comm framework |
|
|
| draft-han-rtgwg-agent-gateway-intercomm-framework | Gateway intercomm |
|
|
| draft-li-dmsc-inf-architecture | DMSC infrastructure architecture |
|
|
| draft-li-dmsc-macp | Multi-agent collaboration protocol suite |
|
|
| draft-liu-dmsc-acps-arc | Agent collaboration protocols arch |
|
|
| draft-yang-dmsc-ioa-task-protocol | IoA task protocol |
|
|
| draft-yang-ioa-protocol | IoA protocol |
|
|
|
|
**Gap:** No draft addresses dynamic trust establishment between gateways, or how to handle conflicting semantic schemas across ecosystems.
|
|
|
|
### 3. Agent Discovery (6 drafts)
|
|
|
|
All trying to solve: how do you find and invoke an AI agent?
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-cui-ai-agent-discovery-invocation | Discovery + invocation protocol |
|
|
| draft-gaikwad-woa | Web of Agents |
|
|
| draft-mozley-aidiscovery | AI Discovery problem statement |
|
|
| draft-mozleywilliams-dnsop-bandaid | DNS-based agent discovery |
|
|
| draft-narajala-ans | Agent Name Service (DNS-based) |
|
|
| draft-pioli-agent-discovery | Agent Registration & Discovery Protocol |
|
|
|
|
**Gap:** None adequately addresses discovery across trust boundaries or privacy-preserving discovery (finding agents without revealing your intent).
|
|
|
|
### 4. Intent-Based Agent Routing (5 drafts)
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-ainp-protocol | AI-native semantic protocol |
|
|
| draft-li-semantic-routing-architecture | Semantic routing architecture |
|
|
| draft-sun-zhang-iaip | Intent-based agent interconnection |
|
|
| draft-sz-dmsc-iaip | Same, different WG |
|
|
| draft-zeng-opsawg-llm-netconf-gap | LLM-driven netconf gap analysis |
|
|
|
|
**Gap:** No draft defines how to verify that an agent's stated intent matches its actual behavior.
|
|
|
|
### 5. 6G Agent Requirements (6 drafts)
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-du-ai-agent-communication-6g-aspect | 6G agent comm requirements |
|
|
| draft-hw-ai-agent-6g | Agent protocol requirements for 6G |
|
|
| draft-liu-dmsc-gw-requirements | Gateway requirements |
|
|
| draft-mao-rtgwg-apn-framework-for-ioa | APN framework for IoA |
|
|
| draft-rosenberg-aiproto-framework | General agent protocol framework |
|
|
| draft-tong-network-agent-use-cases-in-6g | 6G use cases |
|
|
|
|
**Gap:** Heavy on requirements, light on actual protocol mechanisms. Most are wish lists rather than specifications.
|
|
|
|
### 6. Agent Identity via SCIM/Registry (6 drafts)
|
|
|
|
| Draft | Approach |
|
|
|-------|----------|
|
|
| draft-abbey-scim-agent-extension | SCIM extension for agents |
|
|
| draft-liang-agentdns | Root domain naming for agents |
|
|
| draft-scim-agent-extension | Same as abbey (co-submitted) |
|
|
| draft-wahl-scim-agent-schema | SCIM agentic identity schema |
|
|
| draft-zheng-agent-identity-management | Agent identity in IoA |
|
|
| draft-zheng-dispatch-agent-identity-management | Same, different WG |
|
|
|
|
**Gap:** No draft addresses identity lifecycle — what happens when an agent is retired, compromised, or forked?
|
|
|
|
---
|
|
|
|
## Underserved Areas (potential gaps to fill)
|
|
|
|
Based on what's NOT well covered:
|
|
|
|
1. **Agent accountability & audit trails** — Only draft-aylward-daap-v2 and draft-birkholz-verifiable-agent-conversations tackle this seriously
|
|
2. **Token/cost efficiency in agent protocols** — Only draft-chang-agent-token-efficient addresses token bloat
|
|
3. **Agent safety constraints at the protocol level** — AI safety/alignment has highest avg score (3.8) but fewest drafts (36)
|
|
4. **Cross-trust-boundary agent delegation** — Chained auth across org boundaries barely addressed
|
|
5. **Privacy-preserving agent discovery** — No draft covers this
|
|
6. **Agent behavior verification** — Intent vs. actual behavior mismatch detection
|
|
7. **Agent identity lifecycle** — Retirement, revocation, compromise recovery
|